@inproceedings{evaluationexamplesarenotequally,
  title = {Evaluation Examples are not Equally Informative: How should that change NLP Leaderboards?},
  author = {Pedro Rodriguez and Joe Barrow and Alexander Miserlis Hoyle and John P. Lalor and Robin Jia and Jordan Boyd-Graber},
  year = {2021},
  booktitle = {ACL 2021 5},
  url = {https://aclanthology.org/2021.acl-long.346},
}